Training in progress, epoch 0

config.json

@@ -1,23 +1,20 @@
 {
   "architectures": [
-    "T5LAForConditionalGeneration"
+    "T5LaForConditionalGeneration"
   ],
-  "attribute_map": {
-    "hidden_size": "d_model",
-    "num_hidden_layers": "num_layers"
-  },
   "classifier_dropout": 0.0,
   "d_ff": 2816,
   "d_kv": 64,
   "d_model": 1024,
   "decoder_start_token_id": 0,
-  "dense_act_fn": "relu",
+  "dense_act_fn": "gelu_new",
   "dropout_rate": 0.1,
+  "dtype": "float32",
   "eos_token_id": 1,
-  "feed_forward_proj": "gated-gelu",
+  "feed_forward_proj": "gated-gelu_new",
   "initializer_factor": 1.0,
   "is_encoder_decoder": true,
-  "is_gated_act": false,
+  "is_gated_act": true,
   "layer_norm_epsilon": 1e-06,
   "lookahead_size": 2,
   "lookahead_type": "laa",
@@ -28,8 +25,7 @@
   "pad_token_id": 0,
   "relative_attention_max_distance": 128,
   "relative_attention_num_buckets": 32,
-  "torch_dtype": "float32",
-  "transformers_version": "4.49.0.dev0",
+  "transformers_version": "4.56.0.dev0",
   "use_cache": true,
   "vocab_size": 32128
 }

runs/Sep03_11-24-48_gpu21.viking2.yor.alces.network/events.out.tfevents.1756895141.gpu21.viking2.yor.alces.network.1766471.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55839b8db7e5ea65dcdcae2270cab333111b44b3456c08806aacb6c246ad72b3
+size 5128

runs/Sep03_11-28-22_gpu21.viking2.yor.alces.network/events.out.tfevents.1756895347.gpu21.viking2.yor.alces.network.1767250.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f4e8310841cb0ea7fa663710ad62afabaa32194fc49e0e2587bc721e212b73b
+size 7559

spiece.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d60acb128cf7b7f2536e8f38a5b18a05535c9e14c7a355904270e15b0945ea86
+size 791656

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:088b9d391912f110686ce196a363b821cc445f35b1cb5246bd30cab09fc85ead
-size 5432
+oid sha256:ca44cbce5a09fae8b7b0f876e9b92785393d26c08fe74a7a9149c181d217958a
+size 5905